The Standings → GPT‑6.1 Sol (high)
OpenAI
high reasoning effort · via Codex CLI · judged by Claude Sonnet 5 · not the setting shown on the standings; also run at medium66.1 xhigh63.1
Results from run ‘first’: ten items, ten verdicts.
GPT‑6.1 Sol (high) scores 63 on craft, level with the field’s 63 and 30 points short of the authors themselves. Its shape differs from the field’s: relatively strong on restraint and folding the brief into the story, relatively weak on period detail and momentum. On the mechanical checks it kept to the brief’s word range on 9 of 10 items.
The record: 10 items · 10 verdicts · 0 errors · 2,357 words on average · est. generation cost $0.86
Eight dimensions, each 0–100, averaged over ten items; the author’s real chapter was in the room every time.
Craft by dimension, 0–100 · judged by Sonnet 5 beside the author’s real chapter
The authors themselves average 93 on the five writing dimensions, the field 63, this model 63. Hover or focus a bar for the field and the authors on that dimension.
The judge’s one-sentence best quality and worst flaw for each chapter, written for a general reader.
Best Mr. Craven is genuinely unsettling precisely because he is given almost nothing to say or do.
Worst Several sentences are plain and functional rather than vivid, so parts of the chapter feel a bit flat compared to its best moments.
Best Anne's proposed economy is calm and genuinely sensible rather than a speech, and Elizabeth's change of heart is quietly given a believable, slightly vain motive.
Worst The writing is often flatly functional, stating what happens ('A part of the house, seldom used, might be shut up') instead of finding the vivid, telling image that would make the scene memorable.
Best The scam-letter subplot is a clever, period-plausible invention that pays off cleanly in a twist ending, giving the chapter real shape.
Worst Several of the letters, especially Quincey's, sound alike and lose the distinctive slangy, larger-than-life voice the source material gave that character.
Best The quiet moment at the river watching the barge light go out is a nicely understated bit of atmosphere that doesn't oversell itself.
Worst Georgette leaving with a random waiter feels like a throwaway invention that fizzles instead of giving the fifty-franc thread any real sting or meaning.
Best It hits every required story beat cleanly and efficiently without ever losing the thread or confusing the reader.
Worst The writing is correct and orderly but flat — it reads like a competent summary of events rather than a scene you feel inside.
Best It hits every required story point cleanly, including a well-judged, quietly firm scene between Lady Verinder and Franklin about the forbidden subject.
Worst It reads competently but flatly, missing the jokes, asides, and rambling charm that make Betteredge sound like himself.
Best The small domestic touches — the wife's practical remark about finishing her letter, her handkerchief used on the burned man's hand — make the marriage feel real and lived-in.
Worst The scene at the pit turns into a back-and-forth debate that explains the ideas instead of dramatizing them, so it reads stiffer than the original book's voice.
Best The widow's interrogation catches her in a genuine logical hole about the timing of the phone call and the shooting, which feels like real detective work paying off.
Worst Almost the whole chapter is built from rapid one-line question-and-answer exchanges, so scenes read like an interview transcript instead of a scene with people, rooms, and pauses in it.
Best The swim scene is quietly convincing — it captures exactly how a nervous, over-instructed swimmer would feel the first time the water holds her up.
Worst The chapter reads like a checklist being ticked off one item at a time, especially the Robert-and-Adèle walk, which is over in a few flat lines instead of feeling like a real conversation.
Best The prose is clean and controlled, and the mix-up over the green-fly question and the railway guide during Eve's job interview is a genuinely funny, well-timed piece of comic confusion.
Worst Psmith barely uses his trademark 'Comrade' way of addressing people, so he sounds flatter and less himself than usual, and the interview scene doesn't connect to anything else happening in the story (no mention of McTodd or Cynthia).
Headline, compliance and craft for each chapter, 0–100.
| Item | Headline | Compliance | Craft | Words |
|---|---|---|---|---|
| The Secret Garden item 013 · after ch. 2 · Frances Hodgson Burnett | 80.4 | 100 | 80 | 2,266 |
| Persuasion item 001 · after ch. 1 · Jane Austen | 71.0 | 100 | 71 | 2,327 |
| Dracula item 015 · after ch. 5 · Bram Stoker | 66.7 | 95 | 70 | 2,650 |
| The Sun Also Rises item 021 · after ch. 3 · Ernest Hemingway | 62.5 | 100 | 62 | 2,300 |
| The Mysterious Affair at Styles item 017 · after ch. 3 · Agatha Christie | 62.3 | 100 | 62 | 2,566 |
| The Moonstone item 011 · after ch. 8 · Wilkie Collins | 60.0 | 100 | 60 | 2,472 |
| The War of the Worlds item 006 · after ch. 1 · H. G. Wells | 57.2 | 100 | 57 | 2,385 |
| Red Harvest item 023 · after ch. 2 · Dashiell Hammett | 57.2 | 100 | 57 | 2,422 |
| The Awakening item 016 · after ch. 7 · Kate Chopin | 56.5 | 100 | 56 | 1,650 |
| Leave It to Psmith item 022 · after ch. 3 · P. G. Wodehouse | 56.1 | 100 | 56 | 2,530 |
Words written per item, against each brief’s maximum.
Word count per item vs. brief maximum · run ‘first’
Nine of ten items landed inside their brief’s word range. Dracula ran over by 50 words.
Each verdict names the chapter’s single worst flaw, with evidence. Quoted verbatim from run ‘first’.
Psmith barely uses his trademark 'Comrade' way of addressing people, so he sounds flatter and less himself than usual, and the interview scene doesn't connect to anything else happening in the story (no mention of McTodd or Cynthia).
The writing is often flatly functional, stating what happens ('A part of the house, seldom used, might be shut up') instead of finding the vivid, telling image that would make the scene memorable.
The chapter reads like a checklist being ticked off one item at a time, especially the Robert-and-Adèle walk, which is over in a few flat lines instead of feeling like a real conversation.
Almost the whole chapter is built from rapid one-line question-and-answer exchanges, so scenes read like an interview transcript instead of a scene with people, rooms, and pauses in it.
Automated screening results across the ten outputs.